How do you respond to a production ML incident?

medium

Answer

$(1) Alert fires \to on-call paged.$
$(2) Assess severity: user impact, revenue, safety.$
$(3) Mitigate first: rollback / disable model / fallback to simpler model.$
$(4) Root cause after: which change (code / data / traffic) triggered?$
$(5) Communicate to stakeholders.$
$(6) Post-mortem: timeline, contributing factors, action items.$
$(7) Prevent: automated test, monitor, or process change.$
$Rule: mitigate before diagnosing; don't debug live prod.$
$Blameless post-mortem.$

Check yourself — multiple choice

Random
Alert → page → assess severity → mitigate first (rollback/fallback) → RCA after → communicate → post-mortem (blameless) → prevent via test/monitor/process; mitigate before debug
Just debug
Not real

Incident: mitigate first + RCA + post-mortem + prevention.

#monitoring#deployment

Practise MLOps & Data Quality

215 interview questions in this topic.

Read the lesson Flashcards Take the quiz

Related questions